APPLICATION NOTE

FT-IR and NIR Spectroscopy

Author:
Ian Robertson

PerkinElmer, Inc.
Shelton, CT

Detection of Adulteration
of Glycerol with Diethylene Introduction
Glycerol is a popular alcohol used
Glycol by Infrared Spectroscopy within the food, pharmaceutical, and
personal care industry. Glycerol (or
glycerine) is listed in over 2,000
products, such as toothpaste, mouthwash, and pain relief medication due to its desirable
properties to act as a humectant, preservative, and lubricant. It is prepared from fats and oils
and also as a by-product from biofuel production. Demand for glycerol is expected to
increase by 6.3% to $2.1 billion within the next four years, with North America and Western
Europe being the largest suppliers, accounting for 64% of the market.
There have been a significant number of glycerol adulteration cases found by various
regulatory services around the world. Diethylene glycol (DEG) is the most common adulterant
reported, as the physical properties are similar and it is three times cheaper than glycerol. This
fraudulent activity has caused fatalities due to the toxic nature of DEG. In 2006, cough syrups
in Panama were manufactured with DEG instead of glycerol, resulting in hundreds of
reported deaths. Furthermore, cases of contaminated toothpastes (1% - 8% DEG) have been
reported across Europe, America, and Australia. This resulted in the recall of the products and
stricter regulations by various health organizations on the testing and use of glycerol. As a
result, the FDA stipulates a safety limit of 0.1% DEG by weight in products and propose
testing of the raw materials.
Traditionally, thin layer chromatography (TLC) and gas chromatography (GC) have been
adopted to detect DEG, but this process can be time consuming. Infrared (IR) spectroscopy is
a fast, reliable, and low-cost alternative with minimal volume of samples required. This
application note describes the utilization of Compare and Adulterant Screen within Spectrum
10 (and Spectrum 10 ES for regulated environments) to enable a fast and accurate
verification of samples or to confirm the presence and concentration of the adulterant DEG
down to sub per cent levels.
 Experiment
Before a raw material can be used in a process, tests must be
performed to ensure that it is the correct material and of the
desired grade or purity. Infrared (IR) and near-infrared (NIR)
spectroscopy are often the first or only analytical technique
adopted for this purpose. Companies will have collected reference
spectra of standards (often Pharmacopoeia standards) for all of
their raw materials and will compare the spectrum of the incoming
material against those reference standard spectra. The Spectrum
software uses the patented Compare algorithm for this purpose.
A perfect match between the unknown and the reference spectra
will yield a result of 1.00 with the Pass/Fail threshold normally set
to 0.98. Data has been collected using a Spectrum Two FT-IR Figure 1. Compare result for a test sample of glycerol
Spectrometer equipped with a UATR (diamond ATR) sampling
accessory. Spectra were measured at 4 cm-1 resolution with a data
collection time of about 30 seconds. The result shown in Figure 1
is for a test sample of glycerol where the sample passes the test
with a correlation value greater than the Pass/Fail threshold.
A sample of a potential adulterant, diethylene glycol (DEG), yields
the result shown in Figure 2 where the material correctly fails the
Compare against the glycerol standard.
The infrared spectrum and the Compare result will prevent cases
where the material has been incorrectly labelled as the wrong
material or cases where gross contamination or adulteration
has occurred.
In cases where a material fails the Compare test, the material Figure 2. Compare result for a test sample of diethylene glycol (DEG) vs. glycerol standard
would normally be subjected to further testing by additional
instrumental techniques or wet chemistry. However, PerkinElmer Table 1. Compare results against a glycerol standard
has invented an algorithm for detection of adulterants available Sample Name Best Hit Correlation Pass/Fail
within the Spectrum software in cases where adulteration or
Glycerol C:\pel_data\spectra\ATR\Glycerol.sp 0.999428 Pass
contamination is suspected. The same spectrum collected for the
Diethylene Glycol C:\pel_data\spectra\ATR\Glycerol.sp 0.307619 Fail
Compare test can be used with the Adulterant Screen routine to
test to see if the software can detect any adulterants.
To create an Adulterant Screen for the material, the Spectrum 10
software is used to collect your IR or NIR data representing “good
examples” spectra for your material. This will consist of multiple
spectra representing many sources of variation of your product,
such as different batches, sources, grades of material, etc. These
spectra define the boundaries of the acceptable product. You will
then collect the spectra of all known adulterants as pure materials.
These spectra define your Adulterant Spectra. In this example of
the detection of diethylene glycol in glycerol, we have collected
10 spectra of glycerol to define our acceptable product and one
spectrum of pure diethylene glycol as the adulterant.
Using the same spectrum that we used for our unknown sample
with the Compare algorithm we could test the sample for Figure 3. Adulterant Screen result for a test sample of glycerol
adulteration. The results for the sample previously identified as
glycerol are shown in Figure 3.
The plot shows the difference between the calculated model for
glycerol and the unknown spectrum. The glycerol sample is
reported as containing no adulterants, passing the test with very
small spectral residual between the Adulterant Screen model and
the unknown spectrum.

2
A series of mixtures were prepared by mixing the diethylene glycol
and glycerol in order to test the Adulterant Screen method. Results
are shown for these test samples in Table 2. An example for one of
these test samples is shown in Figure 4.
Including the adulterant in the model greatly reduces the residual
and clearly shows that the adulterant is present. Adulterant Screen
has also given an estimated concentration level for this adulterant,
without the requirement for preparing accurate standards and
measuring their spectra. The software will also estimate the
detection limit of the adulterant, in this case approximately 0.2%.
This makes this mid-infrared measurement perfectly suited to
screening glycerol samples for the presence of diethylene glycol.
Further adulterants could be added to the method in the future by
simply adding in the spectrum of the pure adulterant. If more Figure 4. Adulterant Screen residuals for a test sample of glycerol spiked with 17%
DEG (Red: excluding adulterant, Green: including adulterant)
accurate quantitative measurements are required then the
Spectrum Quant software is capable of generating quantitative Table 2. Calculated levels of DEG adulterant for a series of spiked samples
calibrations for the measurement of diethylene glycol in glycerol
Concentration of Adulterant Screen Level of Adulterant
using infrared data. However, calibrations would need to be Diethylene Glycol Pass/Fail Reported
generated for all potential adulterants. Adulterant Screen removes 1.08% Fail 0.0101
this requirement, and, in this case, the Adulterant Screen 2.20% Fail 0.0203
quantitative estimate of the level of adulterant is accurate. 5.31% Fail 0.0519
In an ideal scenario, these measurements would be performed 10.29% Fail 0.1031
close to the point where the raw materials are received or where 14.96% Fail 0.1488
they are to be used in the process. The measurement should be 20.36% Fail 0.2043
performed by a trained operator, not necessarily the skilled 25.56% Fail 0.2606
chemist. These Adulterant Screen methods can be deployed 30.17% Fai 0.3040
using the Spectrum 10 or Spectrum 10 macros with Adulterant 40.19% Fail 0.4028
Screen software, utilizing a simple user interface, following a 50.25% Fail 0.5143
prescribed method or SOP on the screen. In addition, results can 0% Pass -
be exported to the TIBCO Spotfire® for enhanced visualization
and data analysis.

Summary
Deployment of the Spectrum Two with the UATR sampling module
and the Spectrum 10 software enables fast and accurate testing of
the glycerol raw material. The Compare algorithm achieves
identification of the correct raw material. The Adulterant Screen
method is very effective in cases when it is required to screen
against suspected economic adulterants. This simple and effective
method has the ability to detect the adulteration by DEG in
glycerol raw ingredient, down to 0.2% (w/w). An additional
advantage is that Adulterant Screen allows new adulterants to be
added at any time, and, along with the implementation as a
Spectrum macro, it enables easier, simplified use. Spectrum 10 ES
provides all of the security features required for deployment in Figure 5. TIBCO Spotfire® software displaying multiple results for Adulterant Screen
regulated environments.

PerkinElmer, Inc.
940 Winter Street
Waltham, MA 02451 USA
P: (800) 762-4000 or
(+1) 203-925-4602
www.perkinelmer.com

For a complete listing of our global offices, visit www.perkinelmer.com/ContactUs

Copyright ©2014, PerkinElmer, Inc. All rights reserved. PerkinElmer® is a registered trademark of PerkinElmer, Inc. All other trademarks are the property of their respective owners.

011876_01